Livestock Production Science 75 (2002) 209–218www.elsevier.com/ locate / livprodsci

Short communication

Effect of level of inclusion and method of presentation of asingle distillery by-product on the processes of ingestion of

concentrate feeds by horses

*J. HillFaculty of Applied Science and Technology, Department of Agriculture and Rural Management, Writtle College, Chelmsford,

Essex CM1 3RR, UK

Received 26 November 2000; received in revised form 27 June 2001; accepted 19 August 2001

Abstract

The eating behaviour of horses offered wheat distiller grains at varying levels in the concentrate was examined in twoexperiments. The objective of the study was to evaluate the effect of rate of inclusion and presentation (dry vs. soaked) ofwheat distiller’s grains, and the responses in behaviour and processes of intake. In both experiments, the rates of inclusionwere 1.00:0, 0.75:0.25, 0.50:0.50, 0.25:0.75 and 0:1.00 for wheat distiller’s grains /concentrate, respectively. The variousmixtures of concentrate were offered to 10 horses according to two 5 3 5 Latin squares for 10 min and the rate of eating andprocesses of ingestion were assessed. When offered at a rate of 0.75 in the dry matter and not soaked, there was a significantreduction (P , 0.01) in the rate of ingestion of DM and chews per kg DM (P , 0.01). However, the rate of ingestion ofsoaked mixed concentrate declined rapidly (P , 0.01) when the rate of inclusion was greater than 0.5 of the dry matter. Thebehavioural responses of the animals were commensurate to the observations of rate of eating in the case of non soakedconcentrate. When 0.75 of the concentrate DM was replaced with wheat distiller’s grains, a significant increase (P , 0.05) inthe incidence of shorter than expected feeding bouts was observed. However if feeding was terminated, reversion of feedingoccurred with a mean frequency of 0.767. The rate of cessation of feeding was also increased significantly when 0.75 or allof the concentrate DM was replaced with the distillery by-product. If the concentrate was soaked before offering, there wasan increase (P , 0.05) in the number of feeding bouts when 0.25 of the concentrate was replaced with wheat distiller’sgrains. However the processes of ingestion were not affected until 0.5 of the concentrate DM was replaced. Cessation offeeding was affected significantly by the previous behaviour (P , 0.001) especially when the animal interacts with theenvironment and resumption of feeding was reduced significantly (P , 0.01) as the rate of inclusion of wheat distiller’sgrains increased. Wheat distiller’s grains can be used as a substitute to other energy and protein feeding-stuffs in the ration ofthe horse. However the inclusion rate depends on the method of presentation of the feed to the animal. Soaking of theconcentrate prior to feeding allowed a reduced level of distillery by-product to be incorporated into the ration. Least costformulation of concentrates for horses should therefore consider the nutritive composition of the feeding stuff, the potentialof the novel feeding-stuff to disrupt behaviour during feeding and the cost of the formulation. 2002 Elsevier ScienceB.V. All rights reserved.

Keywords: Distillery by-products; Horse; Voluntary intake; Behaviour

*Tel.: 144-1245-424-200; fax: 144-1245-420-456.E-mail address: jh@writtle.ac.uk (J. Hill).

0301-6226/02/$ – see front matter 2002 Elsevier Science B.V. All rights reserved.PI I : S0301-6226( 01 )00304-9

210 J. Hill / Livestock Production Science 75 (2002) 209 –218

1. Introduction that could lead to a dysfunction in the process ofingestion or illness (Houpt, 1990) and these sen-

Distillery by-products are generally high in sitivities may partially explain the commonly ac-metabolisable (approximately 12.5 to 13.0 MJ/kg cepted concept of palatability (‘like’ or ‘dislike’) ofDM) and crude protein (300–340 g crude protein /kg feeds. The principal objectives of the experimentDM). They are utilised extensively as a component reported in this paper were to examine the effects onof ruminant concentrate feeds. However, the rate of process of eating and pattern of feeding of horsesinclusion of distillery by-products is generally re- offered concentrates containing distillery by-productsstricted by the presence of anti-feeding agents that at different levels of inclusion.can either confer a bitterness to the feed or reducethe availability of protein in the rumen and smallintestine (Chamberlain and Wilkinson, 1996). 2. Material and methods

The use of distillery by-products in equine rationsis limited (Leonard et al., 1975). Estimations of 2.1. Experimental designdigestible energy of distillery by-products range from11.5 to 14.2 MJ/kg DM (DLG, 1995) and con- Two experiments were performed. The first ex-centrations of crude protein of greater than 250 g/kg periment examined the effect of addition of dryDM may allow the products to be an alternative to wheat distiller’s grains to the concentrate given tosoya-bean meal or dried skimmed milk powder horses and the second study investigated the effect of(Frape, 1998). Wheat distiller’s grains also contain a feeding to horses the mixtures of concentrates after arelatively high concentration of oil and copper which period of soaking with water for 24 h. In bothmay be an important alternative energy and mineral experiments, 10 horses (mean body weight 544 kg,supply to performance horses (DLG, 1995; Orme et S.E. 23.1 kg) were allocated to two 5 3 5 Latinal., 1997). Equids can however be very sensitive to squares with five horses per square and five concen-the inclusion of novel feeds into the diet. The trate treatments per square. Five ratios of inclusionsensitivity to the novel feed or compound may be a of wheat distiller’s grains were used in the experi-taste-aversion response or a neophobic response ment being 1.00:0, 0.75:0.25, 0.50:0.50, 0.25:0.75(Launchbaugh, 1995). Randall et al. (1978) ex- and 0:1.00 for wheat distiller’s grains /concentrate onamined a range of novel compounds added to the a DM basis, respectively. Horses were offered 1.0 kgdiet of horses. Compounds with acidic or bitter tastes DM of concentrate mixtures according to the se-(for example acetic acid or quinine) were rejected as quence of the Latin square design. No period ofsolutions but when added to solid food, higher adjustment to the feeding stuff occurred and theconcentrations of the compounds were tolerated. concentrate mixtures were offered at 13:00 h. AThese observations have implications for the man- period of 2 days between each treatment occurredagement of the horse in certain circumstances for with no inclusion of distillery by-product. Theexample horses with chronic obstructive pulmonary concentrate mixtures were offered to the animals fordisorder where the animal has to be fed diets that a period of 10 min via a feed tray (0.5 m diameter)have been pre-soaked. placed on the floor of the stable at 13:00 h.

Ellis and Hill (2000) investigated the pattern ofchewing and short-term intake rate of horses offered 2.2. Animal management and feedslucerne feeds (high-temperature or fermented), andobserved an increase in comminution but a decrease The horses were housed individually and beddedin overall rate of ingestion of DM in horses offered on straw. The horses were offered a diet comprisingthe fermented feed compared to the dry feed. Also, 0.7 (7.0 kg DM/day) of medium quality perennialwhen offered in a choice selection experiment rye grass hay (Table 1 for composition) and 0.3 (3.0(Hawkes et al., 1985), the dry lucerne feed was eaten kg DM/day) of concentrate. The chemical com-in preference to the fermented feed. Horses can position of the hay, concentrate and wheat distiller’sdevelop taste aversion learning to avoid certain feeds grains are in Table 1. The concentrate was based on

J. Hill / Livestock Production Science 75 (2002) 209 –218 211

Table 1 I: ingestion sampling (bites, nodding and blow-Chemical composition (g /kg DM unless otherwise stated) of hay ing),and concentrates offered

CT: chewing with tongue manipulation (head upHay Concentrate Wheat position and chewing of feed with pre-molars),

distiller’s C: chewing without manipulation of feed (headgrains

down position, rapid ingestion of feed),DM (g DM/kg) 844 857 852 IE: interaction with environment of the stableNDF 672 344 347

(head shaking, eating bedding, alert response),ADF 365 187 187L: Licking response — animal shows salivationADL 20.3 18.2 14.5

Crude fibre 427 109 76 response at end of feeding sequence,Crude protein 87 129 279 RF: rejection of feed (animal moves away andAsh 92 68 54 shows no interest in eating).Oil (AHEE) 21 52 126Calcium 5.1 4.7 2.2

Events I, CT and C were classified as behaviouralPhosphorus 2.2 3.4 3.8Estimated ME (MJ/kg 8.8 11.7 12.5 types constituting the process of eating and thereforeDM; ruminant) constituted the feeding bout. The behaviour associ-

ated with sampling of feeds (SO) was classified asthe event that initiated feeding. However, reversion

wheat, barley, oats (cereal component 0.65 of DM), of feeding was determined if the animal resumedsoya, sunflower seed and locust bean (protein feeds feeding (behavioural class I) after short periods of0.25 of DM) and nutritionally improved straw and behaviour (25 s) not associated with feeding (IE orlucerne meal (0.10 of DM). The concentrate mix- L).tures were offered either as a dry mixture (Experi-ment 1) or soaked with water (two parts concentrate 2.4. Statistical proceduresDM to one part water w/w) for 24 h prior to feeding(Experiment 2) at 13:00 h. The concentrate feeds In both experiments, analysis of varianceoffered at 07:00 and 17:00 h contained no distillery (ANOVA) was applied to log 10 transformed databy-product. The chemical composition of forage and collated for the rate of ingestion (g DM/min), rate ofconcentrates were analysed according to the methods chewing (number of chews/min), feed intake per biteof Hill and Leaver (1999). (g DM eaten per number of bites recorded as

behavioural events), feed intake per bout (g DM/bout) according to the model:

2.3. MeasurementsYij 5 m 1 bi 1 kj 1 t(t) 1 ´ij

After the period of 10 min of offer of concentrate where Yij is the observation at the intersection of themixtures, the refusals were weighed and dried at ith row and the jth column and bi represents the row1008C for 24 h to determine the rate of ingestion of (period) effect, kj represents the column (animal)concentrate DM. The behaviour of the horses was effect, t(t) represents the treatment effect (repre-recorded continuously using a monochrome CCD sented in each row or column once) and ´ij thecamera (speed of 25 images per s) and a video unaccounted error. In both experiments, the responserecorder. The images were examined by developing to treatment was estimated on 28 error degrees ofan ethogram to analyse the detailed changes in freedom (d.f.), with main effects of treatments (4position of the horse to the feed and the processes of d.f.), periods (8 d.f.) and animals (8 d.f.) beingchewing and eating (continuous sampling of be- considered by ANOVA. The effect of squares washaviour). The data were classified into seven main removed accounting for one degree of freedom.types; Transition matrices were calculated (frequencies

of behavioural changes) from the behavioural ob-SO: sampling feed by olfaction (sniffing), servations according to Haccou and Meelis (1992).

212 J. Hill / Livestock Production Science 75 (2002) 209 –218

Interaction with the environment (IE) and behaviour DM (P,0.01; Table 2). A reduction in intake of DMassociated with termination of feeding (L and RF) per bout of feeding was also observed when all thewere examined for occurrence during the period of concentrate DM was replaced by wheat distiller’soffer. The relationships between mean ingestion bout grains (P,0.01). As the rate of eating declined withlength (duration of time spent in behavioural classifi- increasing level of addition of wheat distiller’scations I, CT and C) and category of preceding grains, there was a reduction in the chews/g DMbehavioural act were examined using Kruskal–Wallis (P,0.01; Table 2). However, the rate of chewing ofANOVA. Proportions of abortive feeding events the different concentrates remained similar in all(animal changes from behaviour I to any behavioural treatments with a mean (of all treatments) rate ofevent not related to intake i.e. C or CT) in total chewing of 55 (S.E. 5.2) chews/min.number of behavioural events, shorter and longerthan expected feeding bouts (bouts exceeding the 3.1.2. Experiment 2: effect of soaking concentrateduration of 1.96 s of mean calculated duration of mixturesfeeding bout), interruptions of bouts by other be- The effect of soaking the concentrate mixtureshaviour and reversion to feeding after interruption of prior to feeding lead to a significant reduction in thefeeding caused by other behavioural events were rate of ingestion of feed DM when 0.50 of thecalculated according to Kalueff (2000). All statistical concentrate DM was replaced by wheat distiller’sanalyses were performed using Unistat Version 4 grains (P,0.001; Table 3). The reduction of chews/(UNISTAT, 1995). g DM (P,0.05) was not however as great as

expected if the results were considered with Experi-ment 1and the rate of chewing was reduced sig-

3. Results nificantly (P,0.05) when all the concentrate wasreplaced with wheat distiller’s grains (Table 3). The

3.1. Feed intake rate of chewing of the different concentrates re-mained similar in all treatments with a mean (of all

3.1.1. Experiment 1: inclusion of dry distillery by- treatments) rate of chewing of 53 (S.E. 6.9) chews/product min except when all the concentrate was replaced by

In Experiment 1, the inclusion of wheat distiller’s wheat distiller’s grains. When the diet was totallygrains into the concentrate component of the diet replaced, chewing rate was reduced to 38 (S.E. 7.1)reduced significantly the rate of ingestion of feed chews/min.

Table 2 Table 3Ingestion characteristics of horses offered varying levels of wheat Ingestion characteristics of horses offered varying levels of wheatdistiller’s grains in the concentrate: concentrate offered as dry distiller’s grains in the concentrate: concentrate offered as wetfeedingstuff (soaked) feedingstuff

Inclusion of wheat Ingestion g DM/ Ingestion rate Inclusion of wheat Ingestion g DM/ Ingestion ratedistiller’s grains rate (g eating (g DM/chew distiller’s grains rate (g eating (g DM/chew(proportion of DM/min) bout during eating (proportion of DM/min) bout during eatingdiet DM) bout) diet DM) bout)

a a a a0 203.1 110.0 3.7 0 177.9 92.9 3.4ab a a a0.25 171.6 110.7 3.2 0.25 155.4 91.2 2.9b a b ab0.5 167.1 100.3 3.0 0.5 92.5 76.2 1.7b a c b0.75 154.2 102.5 2.8 0.75 55.2 62.5 1.1c b c c1.00 101.4 44.3 1.8 1.00 39.2 18.6 1.0

S.E.D. 14.84 11.43 S.E.D. 12.48 10.52a,b,c a,b,cP,0.05 indicates level of significance of between treatment P,0.05 indicates level of significance of between treatment

differences. differences.

J. Hill / Livestock Production Science 75 (2002) 209 –218 213

3.2. Behavioural responses

3.2.1. Experiment 1: inclusion of dry distillery by-product

Table 4When dry wheat distiller’s grains replaced all theTransition matrices of horses (frequency of change in behaviour)

concentrate DM, there was a reduction (P,0.05) in offered varying levels of wheat distiller’s grains in the concen-change from C to CT (0.326 to 0.254 for 0.75 and trate: concentrate offered as dry feedingstufftotal replacement, respectively) and CT to C (0.298

0 Wheat distiller’s grainsbto 0.228 for 0.75 and 1.0 replacement, respectively; SO CT C IE L RF

aTable 4). A concurrent increase in the proportion of SO 0.568 0.355 0.097CT 0.505 0.455 0.040behaviour associated with interaction with the en- ]] ]]C 0.375 0.337 0.288vironment was noted (C to IE; 0.185 to 0.224 for ]]IE 0.195 0.261 0.187 0.357

]]0.75 and total substitution, respectively; Table 4).L 0.140 0.550 0.310

The mean total number of feeding bouts increased RF 0.333 0.667with the inclusion of wheat distiller’s grains into the

0.25 Wheat distiller’s grainsconcentrate when the rate of inclusion was increasedSO CT C IE L RFfrom 0 to 0.50 of the concentrate DM. However the

SO 0.588 0.344 0.068number of feeding bouts then remained relativelyCT 0.444 0.425 0.131

]] ]]constant at even higher rates of inclusion of the C 0.385 0.332 0.069 0.214]]distillery by-product (Table 5). Feeding events shor- IE 0.192 0.235 0.250 0.323]]

L 0.145 0.552 0.303ter than predicted occurred at a rate of 0.397 whenRF 0.345 0.655no distillery by-product was added to the diet and

increased to 0.479 (P,0.05) only when the concen-0.50 Wheat distiller’s grains

trate was replaced completely by wheat distiller’s SO CT C IE L RFgrains. Also, the effect of preceding event (IE, L or SO 0.655 0.224 0.121

CT 0.313 0.445 0.105 0.137RF) on the disruption of the processes of ingestion ]] ]]C 0.333 0.254 0.055 0.358were significant, especially in the development of ]]IE 0.201 0.225 0.255 0.319

]]shorter than normal feeding bouts (P,0.05) whenL 0.185 0.200 0.455 0.160

the diet was replaced totally by wheat distiller’s RF 0.315 0.685grains. A similar pattern was noted with feeding

0.75 Wheat distiller’s grainsbouts longer than predicted, however the incidence,SOCTCIELRFdecline in occurrence and effect of preceding eventSO 0.768 0.145 0.087(IE and RF only) were lower than the observationsCT 0.298 0.552 0.088 0.062

]] ]]made for shorter feeding bouts (Table 5). C 0.326 0.185 0.489]] ]] ]]The number of abortive feeding bouts (I) was very IE 0.168 0.199 0.288 0.345]]

L 0.224 0.125 0.352 0.299low in all treatments, on average accounting for lessRF 0.295 0.705than 0.15 of all feeding bouts and the frequency of

behavioural events not associated with eating inter-1.00 Wheat distiller’s grains

rupting feeding bouts was low and showed little SO CT C IE L RFvariation between treatments (Table 5). Unusually, SO0.7940.0760.130

CT 0.228 0.521 0.227 0.024abortive feeding bouts did not seem to be affected by ]] ]]C 0.254 0.224 0.227 0.295the preceding behavioural act (IE or RF). If feeding ]] ]] ]]IE 0.154 0.214 0.112 0.520

]]was interrupted by another behavioural event, re-L 0.090 0.182 0.075 0.108 0.545

sumption of feeding occurred with a mean frequency RF 0.272 0.728of 0.767. However, when 0.75 of the concentrate

Behavioural frequencies underlined related to important transi-DM was substituted, there was an increase (P,0.05) tions within process of eating.

ain the rate of cessation of feeding (Table 5). Total First behavioural event.breplacement of the concentrate with wheat distiller’s Subsequent behavioural event (Haccou and Meelis, 1992).

214 J. Hill / Livestock Production Science 75 (2002) 209 –218

Table 5Incidence and type of abortive feeding of horses offered varying levels of wheat distiller’s grains in the concentrate: concentrate offered asdry feedingstuff

Inclusion of wheat distiller’s grains into the concentrate

0 0.25 0.50 0.75 1.00

Total no. of feeding cycles 5.9 6.2 6.3 6.8 7.5per eating bout

Proportion of shorter than 0.479 0.472 0.445 0.461 0.397expected events

Proportion of longer than 0.117 0.165 0.184 0.166 0.205expected events

Proportion of abortive 0.137 0.146 0.155 0.152 0.167events

Proportion of interruptions 0.127 0.129 0.138 0.112 0.155with behaviour other than eating

Post interruption event reversion 0.896 0.822 0.755 0.727 0.667to feeding (frequency)

grain further increased the rate of cessation of reduction in incidence (P,0.01) when 0.75 of thefeeding and was sensitive to the type of preceding concentrate was replaced with the distillery by-prod-act (P,0.01). uct. As in Experiment 1, the type of preceding event

(IE, L or RF) affected significantly (P,0.001) the3.2.2. Experiment 2: effect of soaking concentrate duration of the shorter than normal feeding bout,mixtures however the effect was noted when 0.50 of the

The transitions from C to CT and CT to C were concentrate was replaced by the distillery by-prod-significantly reduced (P,0.001 and P,0.05, respec- uct. The incidence of longer than expected feedingtively) by the addition of wheat distiller’s grains to bouts was similar to those observed in Experiment 1,the concentrate. However the level at which dysfunc- however there was no marked decline in frequencytion in feeding behaviour occurred differed between (Table 7). However, the effect of the preceding actthe transition sequences. In the case of C to CT the (IE, L or RF) on the development of longer thanreduction in frequency occurred after 0.50 of the diet expected feeding bouts occurred when 0.75 of thewas replaced by distillery by-product, however the concentrate was replaced by wheat distiller’s grains.reverse behavioural sequence declined only when In contrast to Experiment 1, there was a significant0.75 of the diet was substituted (Table 6). An increase (P,0.01) in the frequency of abortive boutsincrease in the frequency of eating behaviour (CT of feeding (I). The increased incidence was firstand C) to interaction with the environment (IE) or noted when 0.25 of the concentrate DM was replacedrejection of feed (RF) was observed, however in with wheat distiller’s grains and continued until 0.75contrast to the eating transitions, the increases were of the concentrate was substituted. The increase inobserved after 0.50 of the concentrate DM was incidence of abortive feeding when greater than 0.75replaced by wheat distiller’s grains (Table 6). of the concentrate DM was replaced with wheat

The total number of feeding bouts increased when distiller’s grains was not significant (0.592 to 0.6210.25 of the concentrate was substituted with wheat for 0.75 and complete substitution, respectively, P.

distiller’s grains, but then a decline in the number of 0.05). The development of an abortive feeding boutbouts was noted when all the concentrate was was also sensitive to the preceding behaviouralsubstituted with distillery by-product (Table 7). The event. The rate of reversion to feeding after interrup-incidence of shorter than expected bout length was tion of a feeding bout was reduced significantlyhigher than in Experiment 1 and there was a (P,0.001) by the inclusion of wheat distiller’s

J. Hill / Livestock Production Science 75 (2002) 209 –218 215

grains, with incidences being as low as approximate-ly 0.24 when complete substitution of the baseconcentrate occurred (Table 7). The effect of thepreceding event (IE and RF only) was again found to

Table 6 be significant (P,0.05) in the rate of resumption ofTransition matrices of horses (frequencies of changes in be-

feeding with modifications of behaviour being notedhaviour) offered varying levels of wheat distiller’s grains in thewhen 0.25 of the concentrate was replaced withconcentrate: concentrate offered as wet (soaked) feedingstuffwheat distiller’s grains.

0 Wheat distiller’s grainsbSO CT C IE L RF

aSO 0.535 0.310 0.065 0.090CT 0.544 0.427 0.029 4. Discussion]] ]]C 0.324 0.389 0.287

]]IE 0.105 0.227 0.137 0.531

]] Least cost formulation of concentrate diets forL 0.110 0.640 0.250

herbivores can lead to the incorporation of plantRF 0.368 0.632products not usually offered to the animal. Distillery

0.25 Wheat distiller’s grains by-products are the residual organic matter resultingSO CT C IE L RF from the fermentation of small grain cereals by the

SO 0.577 0.264 0.159 brewing and distillation industry (Leonard et al.,CT 0.512 0.374 0.102 0.012

]] ]] 1975). They have been viewed by the ruminant andC 0.365 0.388 0.025 0.234]] monogastric feed sector as an important source ofIE 0.120 0.269 0.144 0.467]]

L 0.137 0.105 0.450 0.308 carbohydrate, protein and lipid (Lonsdale, 1989).RF 0.400 0.600 Incorporation of these products into equine diets may

however be limited to certain sectors or types of feed0.50 Wheat distiller’s grains

even though the chemical compositions are similarSO CT C IE L RF(Table 1) to the more traditional high protein–oilSO 0.700 0.200 0.100

CT 0.493 0.266 0.105 0.137 supplements. The substitution of more expensive]] ]]C 0.320 0.277 0.403 protein and oil sources (for instance soya bean meal)]]IE 0.226 0.118 0.127 0.529

]] may be feasible with wheat distiller’s grains (Ott etL 0.255 0.095 0.424 0.226

al., 1979; NRC, 1989; Cuddeford, 1999). RecentRF 0.441 0.559interest in the utilisation of oil by the performance

0.75 Wheat distiller’s grains horse could lead to an increase in the incorporationSO CT C IE L RF of distillery by-products into equine diets. The

SO 0.717 0.221 0.062 benefits of including oil supplements in performanceCT 0.217 0.553 0.104 0.126

]] ]] horse diets are twofold; the increased energy densityC 0.322 0.425 0.253]] of the diet (Hintz, 1974) and the possible glucoseIE 0.109 0.127 0.212 0.552]]

L 0.100 0.105 0.521 0.274 sparing effect of elevated mobilisation of free fattyRF 0.273 0.727 acids (Eaton et al., 1995; Orme et al., 1997; Harris,

1997).1.00 Wheat distiller’s grains

The rate of eating or intake processes (bite weight,SO CT C IE L RFchews per bite or chews per kg DM) of horsesSO 0.695 0.055 0.110 0.140

CT 0.167 0.617 0.107 0.109 offered forages or concentrates have not been mea-]] ]]C 0.127 0.239 0.197 0.437 sured extensively (Vernet et al., 1995). The rates of]]IE 0.100 0.153 0.207 0.540

]] ingestion of concentrate feeds offered to horses inL 0.126 0.125 0.276 0.473

the experiments reported in this paper were compar-RF 0.426 0.574able to ruminants (Tables 2 and 3 and Vernet et al.,

Behavioural frequencies underlined related to important transi-1995; Grenet, 1989; Webster, 1979; McGraham,tions within process of eating.

a 1964). However the rate of ingestion of DM andFirst behavioural event.b Subsequent behavioural event (Haccou and Meelis, 1992). processes of eating have been shown to be partially

216 J. Hill / Livestock Production Science 75 (2002) 209 –218

Table 7Incidence and type of abortive feeding of horses offered varying levels of wheat distiller’s grains in the concentrate: Concentrate offered aswet (soaked) feedingstuff

Inclusion of wheat distiller’s grains into the concentrate

0 0.25 0.50 0.75 1.00

Total no. of feeding cycles 6.1 8.5 9.3 10.4 15.3per eating bout

Proportion of shorter than 0.508 0.553 0.575 0.281 0.264expected events

Proportion of longer than 0.230 0.224 0.260 0.203 0.151expected events

Proportion of abortive events 0.016 0.211 0.388 0.592 0.621

Proportion of interruptions with 0.131 0.188 0.247 0.484 0.491behaviour other than eating

Post interruption event reversion 0.750 0.533 0.375 0.313 0.235to feeding (frequency)

dependent on the presentation of the feed (Tables 2 1998). Equids are colonic fermenters and thereforeand 3) and the level of inclusion of the feed. The rate more efficient than ruminants in capture of energyof ingestion of concentrates in general was similar to substrates, especially when diets of very low or verythose observed by Vermorel and Mormede (1991), high nutritive value are offered (Van Soest, 1994;however when the addition of distillery by-product Russell and Gahr, 1999). The rate of ingestion ofexceeded a certain level (0.75 in dry feeding system DM of dry feed is higher than that of wet feedand 0.5 in wet feeding system), a reduction in the (Tables 2 and 3) reflecting a difference in bio-rate of ingestion was noted. It is also important to mechanical properties of wet and dry feed as well asnote that there may be differences in the rate of the modification of behaviour of feeding (Tables 4ingestion during the feeding bout (von Meyer et al., and 6; Lachica et al., 1995). Vernet et al. (1995)1975; Vermorel and Mormede, 1991), however the suggested that the type, structure, chemical com-present study did not address this issue. Horses are position and digestibility of the feed affects thenot, by nature, meal or bout feeders but so-called ingestion rate as time is required for the animal to‘trickle’ feeders and therefore human intervention grip the feed, chew and make the bolus. Thehas attempted to alter the pattern of feeding. Man- magnitude in decline of rate of ingestion of concen-agement of the horse in the stable environment or for trate DM depended on the presentation of the feed.athletic performance has indirectly altered the feed- Rates of addition of 0.75 (dry) and 0.50 (wet)ing behaviour of the horse. Human intervention has reduced the rate of ingestion of feed DM. However,attempted to change the animal to a meal bout the processes involved in the reduction of rate offeeding strategy, by feeding discrete concentrate ingestion are more complex than the precedingmeals twice or three times daily while offering hay at statement suggests and may reflect modifications ofa restricted level of feeding to maintain fibre intake the process of eating without a change in the rate theprimarily (Frape, 1998). Bout feeding strategies have food is processed (Table 4 and 6; Houpt, 1990;the advantage in providing exact levels of digestible Albright, 1993; Launchbaugh, 1995).energy and amino acids; however, do they optimise Few papers have examined in detail the behav-the utilisation of the nutrients? ioural processes of eating of concentrate feeds by

The process of ingestion of feed over a long horses (Randall et al., 1978; Schryver et al., 1987;duration of time (‘trickle-feeding’) reflects the Ellis et al., 2000; Ellis and Hill, 2000). No evidenceanatomy and physiology of the digestive tract (Frape, of neophobia was observed in the behavioural pat-

J. Hill / Livestock Production Science 75 (2002) 209 –218 217

terns of ingestion of the horses used in the experi- facilities and management of animals at Writtlements. If neophobia is classified by periods of low or College.rapidly accelerating intake of feed, even at thehighest rate of inclusion of the dry distillery by-product, the duration and pattern of feeding was not Referencesaltered. The pattern of intake was disrupted morefrequently by other non-ingestive behavioural pat- Albright, J.L., 1993. Feeding behaviour in dairy cattle. J. Dairyterns (Tables 4–7) but little variation in the bout Sci. 76, 485–493.

Chamberlain, A.T., Wilkinson, J.M., 1996. Feeding the Dairylength was observed. When the concentrate wasCow. Chalcombe, Lincoln, UK.soaked, there may be circumstantial evidence of

Cuddeford, D., 1999. Feeding systems for horses. In: Theodorou,neophobia. The inclusion rate of distillery by-prod- M.K., France, J. (Eds.), Feeding Systems and Feed Evaluationuct, the incidence of shorter than expected feeding Models. CAB International, Wallingford, UK, pp. 239–273.cycles and the strong influence of preceding act on Deutsche Landwirtschafts Gesellschaft, DLG, 1995. Futtewet-

tabellen — Ppferde. 3. Ausgabe DLG, Frankfurt am Main,reversion to feeding could suggest neophobia onlyGermany.when distillery by-product was included as the sole

Eaton, M.D., Hodgson, D.R., Evans, D.L., Bryden, W.L., Rose,concentrate. The current experiments did not ex- R.L., 1995. Effect of a diet containing supplementary fat on theamine the consequences of offering distillery by- capacity for high intensity exercise. Equine Vet. J. 18, 353–products in the diets of horses on the general pattern 356.

Ellis, A.D., Fagence, K., Warr, E.M., Hill, J., 2000. Effect ofand duration of nycthemeral ingestion of the totaldental condition on short-term intake rate and faecal particlediet. Therefore, the implications of incorporation ofsize in adult horses. In: Proceedings of the 51st European

distillery by-products to rations of horses have to be Association of Animal Production, The Hague, NL, p. 375.evaluated further with respect to commercial utilisa- Ellis, A.D., Hill, J., 2000. Intake behaviour of horses offered shorttion of the by-product. chopped high temperature dried lucerne and short chopped

lucerne silage. In: Proceedings of the 3rd International Confer-ence on Methods and Techniques in Behavioural Research,Katholieke Universiteit Nijmegen, NL, pp. 101–103.

Frape, D., 1998. Equine Nutrition and Feeding. Blackwell Science,5. ConclusionsLondon.

Grenet, E., 1989. A comparison of the digestion and reduction inIf horses are to be managed intensively utilising particle size of lucerne hay (Medicago sativa) and Italian

concentrates to provide the majority of digestible ryegrass hay (Lolium italicum) in the ovine digestive tract. Br.energy and protein, the process of formulation must J. Nutr. 62, 493–507.

Haccou, P., Meelis, E., 1992. Statistical Analysis of Behaviouralconsider three factors, the acceptability of the con-Data. Oxford University Press, Oxford.centrate to the animal, the nutritive value and the

Harris, P., 1997. Energy sources and requirements of the exercis-cost. Least cost formulation will lead to the use of ing horse. Annu. Rev. Nutr. 17, 185–210.novel plant products in the concentrate feed either to Hawkes, J., Hedges, M., Daniluk, P., Hintz, H.F., Schryver, H.F.,manipulate the digestible energy supply or to im- 1985. Feed preferences of ponies. Equine Vet. J. 17, 20–22.

Hill, J., Leaver, J.D., 1999. Energy and protein supplementation ofprove the supply of amino acids and protein. How-lactating dairy cows offered urea treated whole-crop wheat asever, refinements in the nutritive value are limited bythe sole forage. Anim. Feed Sci. Technol. 82, 177–193.

the acceptability of the feedingstuff to the animal. If Hintz, H.F., 1974. Nutrition and energy performance. J. Nutr. 124,alterations in the behavioural processes controlling 11–12.ingestion are too great, the feed will have limited Houpt, K.A., 1990. Ingestive behaviour. North Am. Equine Pract.

6, 319–337.value to the equine industry.Kalueff, A.V., 2000. Measuring grooming in stress and comfort.

In: Proceedings of the 3rd International Conference on Meth-ods and Techniques in Behavioural Research, KatholiekeUniversiteit Nijmegen, NL, pp. 178–179.Acknowledgements

Lachica, M., Aguilera, J.F., Prieto, C., 1995. Energy expenditurerelated to the act of eating in Granadian goats given diets of

The author wishes to thank E.V. Davidson for different physical form. Br. J. Nutr. 77, 417–426.initial data collection and T.A. Whitaker for use of Launchbaugh, K.L., 1995. Effects of neophobia and aversions on

218 J. Hill / Livestock Production Science 75 (2002) 209 –218

feed intake: why feedlot cattle sometimes refuse to eat nutriti- Schryver, H.F., Parker, M.T., Daniluk, P.D., 1987. Salt consump-ous feeds. In: Symposium: Intake by Feedlot Cattle. Oklahoma tion and the effect of salt on mineral metabolism in horses.State University, Stillwater, OK, pp. 36–48. Cornell Vet. 77, 122–131.

Leonard, T.M., Baker, J.P., Willard, J., 1975. Influence of distillers UNISTAT, 1995. Unistat version 4 for Windows. UNISTAT Ltd.,feeds on digestion in the equine. J. Anim. Sci. 40, 1086–1092. London.

Lonsdale, C., 1989. Straights — Raw Materials for Animal Feed Van Soest, P.J., 1994. Nutritional Ecology of the Ruminant.Compounders and Farmers. Chalcombe, Marlow, UK. Cornell University Press, Ithaca, NY.

McGraham, N.C., 1964. Energy cost of feeding activities and Vermorel, M., Mormede, P., 1991. Energy cost of eating in ponies.energy expenditure of grazing sheep. Aust. J. Agric. Res. 15, In: Wenk, C., Boessinger, M. (Eds.), Energy Metabolism in969–973. Farm Animals. European Association of Animal Production

National Research Council, 1989. Nutrient Requirements of the No. 58. pp. 437–440.Horse, 5th Edition. National Academy Press, Washington, DC. Vernet, J., Vermorel, M., Martin-Rosset, W., 1995. Energy cost of

Orme, C.E., Harris, R.C., Marlin, D., Hurley, J., 1997. Metabolic eating long hay, straw and pelleted food in sport horses. Anim.adaptation to a fat-supplemented diet by the thoroughbred Sci. 61, 581–588.horse. Br. J. Nutr. 78, 443–455. von Meyer, H., Ahlswede, L., reinhardt, H.J., 1975. Unter-

¨Ott, E.A., Asquith, R.L., Feaster, J.P., Martin, F.G., 1979. In- suchungen uber Freßdauer, Kaufrequenz und Futterzer-¨fluence of protein level and quality on the growth and kleinerung beim Pferd. Dtsch. Tierarztliche Wochenschr. 82,

development of yearling foals. J. Anim. Sci. 49, 620–628. 54–58.Randall, R.P., Schurg, W.A., Church, D.C., 1978. Response of Webster, A.J.F., 1979. Energy cost of digestion and metabolism in

horses to sweet, slaty, sour and bitter solutions. J. Anim. Sci. the gut. In: Ruckebusch, Y., Thivend, P. (Eds.), Proceedings of47, 51–55. 5th International Symposium on Ruminant Physiology. MTP

Russell, R.W., Gahr, S.A., 1999. Glucose availability and associ- Press, Lancaster, pp. 469–483.ated metabolism. In: d’Mello, J.P.F. (Ed.), Farm AnimalMetabolism and Nutrition. CAB International, Wallingford,UK, pp. 121–149.